- Email: [email protected]
Differential in vitro hepatotoxicity of troglitazone and rosiglitazone among cryopreserved human hepatocytes from 37 donors
Chemico-Biological Interactions 142 (2002) 57 /71 www.elsevier.com/locate/chembiont
Differential in vitro hepatotoxicity of troglitazone and rosiglitazone among cryopreserved human hepatocytes from 37 donors Scott Lloyd *, Michael J. Hayden, Yumiko Sakai, Andrew Fackett, Paul M. Silber, Nicola J. Hewitt, Albert P. Li In Vitro Technologies, Inc., 1450 South Rolling Road, Baltimore, MD 21227, USA
Abstract We report here our studies on troglitazone and rosiglitazone cytotoxicity in human hepatocytes isolated from multiple donors to investigate factors responsible for individual differences in sensitivity to the known hepatotoxicity of these antidiabetic drugs. Using cellular adenosine triphosphate (ATP) content as an endpoint, cytotoxicity of both drugs was evaluated in cryopreserved human hepatocytes from 37 donors. We confirmed reports of others that troglitazone was cytotoxic to human hepatocytes using cellular ATP content as an endpoint. In addition, we found that rosiglitazone, although less toxic in the study population, was cytotoxic to hepatocytes in some donors (EC50 B/100 mM). ATP content, 3-[4,5dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) metabolism, depletion of intracellular glutathione, Alamar Blue metabolism, and neutral red uptake were used as endpoints in a single donor study using freshly isolated human hepatocytes. Troglitazone appeared to be more toxic than rosiglitazone by all endpoints. From the demographic data provided to us for each donor, we were able to establish no direct correlation between cytotoxicity (expressed as EC50 values) and age, sex, smoking status, or alcohol consumption. We conclude that troglitazone and rosiglitazone are differentially toxic to human hepatocytes, and that toxicity may be independent of age, sex, tobacco use, and alcohol use. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adenosine triphosphate; Hepatocytes; Troglitazone
Support by In Vitro Technologies, Inc., Baltimore, Maryland. This work was presented, in part, at the second International Symposium on Mechanisms, Models and Predictions of Idiosyncratic Drug Toxicity, Atlantic City, New Jersey, June 11 /13, 2001. * Corresponding author. Tel.: /1-410-455-1242; fax: /1-410-455-1245 E-mail address: [email protected] (S. Lloyd).
0009-2797/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 2 ) 0 0 0 5 4 - 6
58
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
1. Introduction The thiazolidinediones are a relatively new class of drugs used to treat (type 2) non-insulin dependent diabetes mellitus, or NIDDM ([1]). This novel class of antidiabetic agents has proven useful as either a monotherapy, or in combination therapy with insulin, metformin or sulfonylureas in the treatment of NIDDM ([1]). The thiazolidinediones are selective agonists of the peroxisome proliferator-activated receptors (PPARs) and have been shown to lower insulin resistance in both diabetic and non-diabetic individuals ([2,3]). The PPARs, especially PPARg, are cell surface receptors and transcription factors present in adipocytes, skeletal muscle, and liver
Fig. 1. Dose response curves to troglitazone (triangles) and rosiglitazone (squares) for cryopreserved hepatocytes from 37 different donors (panels A through KK).
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
59
Fig. 1 (Continued)
tissue, and are key elements in the regulation and metabolism of glucose in response to insulin ([4]). Many patients suffering from NIDDM lack glycemic control due to non-responsiveness of PPARs, making the use of thiazolidinediones an attractive therapeutic approach. Troglitazone was introduced to the market in 1997 and was the first thiazolidinedione to be approved for use in the treatment of NIDDM ([1]). However, during its first 2 years on the market, a number of cases were reported that associated troglitazone with acute liver toxicity and failure, leading to the death of patients using the drug or necessitating liver transplant in some cases ([5,6]). Subsequent clinical investigations lead to its withdrawal from the UK market in December 1997 and to its withdrawal from the US market in March of 2000 based on data that associated troglitazone with cases of serious idiosyncratic liver toxicity ([7]).
60
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
Rosiglitazone, a relatively non-toxic alternative to troglitazone, was introduced to the market in 1999 ([1]). While the incidence of significant hepatotoxicity caused by rosiglitazone has not been observed in controlled clinical trials, two cases have been reported which implicate rosiglitazone as a causative agent of idiosyncratic drug toxicity (IDT) as well ([8 /10]). Several studies have proposed a variety of mechanisms of troglitazone cytotoxicity, including parent compound toxicity in human hepatocytes, the formation of reactive intermediates and glutathione (GSH) conjugates in human liver microsomes, and both cholestatic and apoptotic potentials in cultured rat hepatocytes ([11,12]). Mechanisms of hepatocellular injury caused by rosiglitazone have not been well studied, likely because the causal relationship between the drug and hepatotoxicity remains uncertain. This study used a panel of cryopreserved human hepatocytes isolated from demographically diverse donors,
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
61
Fig. 1 (Continued)
and freshly isolated hepatocytes from a single donor, to demonstrate dose-dependent troglitazone toxicity and to compare individual sensitivities to troglitazone and rosiglitazone on the basis of age, sex, alcohol or tobacco use.
2. Materials and methods 2.1. Incubation of hepatocytes with troglitazone and rosiglitazone Cryopreserved hepatocytes from 37 individual donors, along with freshly isolated hepatocytes from a single donor, were used in this study. The cells were isolated,
62
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
cryopreserved, and thawed using previously published procedures ([13 /15]). Troglitazone and rosiglitazone were obtained from Parke-Davis (Ann Arbor, MI) in powder form and prepared as a 2000 /stock solution in dimethyl sulphoxide (DMSO; Sigma, St. Louis, MO), then diluted to a 2 /concentrated dosing solution in Dulbecco’s Modified Eagle Medium (Sigma). Cytotoxicity assays were initiated by the addition of an equal volume of cryopreserved hepatocyte suspension to the media containing troglitazone or rosiglitazone. The final troglitazone and rosiglitazone concentrations used in the incubations were 10, 20, 50, 100 and 200 mM. The final DMSO concentration was 0.1% (v/v). In a single related experiment, freshly isolated hepatocytes were incubated overnight on collagen (Cohesion, Palo Alto, CA) coated 96 well plates. The plating media was then replaced with incubation
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
63
64
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
media containing 10, 20, 50, 100 or 200 mM troglitazone or rosiglitazone for a 24-h incubation period. Incubations of hepatocytes with drug and control solutions were carried out in a cell culture incubator maintained at 37 8C, 5% CO2 in saturating humidity. The treatment duration was 2 h for cryopreserved hepatocytes.
2.2. Cell viability analysis ATP content analysis was performed using the ATP Lite-M kit from Packard Marketing (Groningen, The Netherlands). Briefly, the hepatocytes were lysed after incubation followed by the addition of luciferin and luciferase. After a 10-min dark adaptation, luciferin luminescence was quantified using a Wallac Victor 2 Microplate Reader (Perkin Elmer, Gaithersburg, MD). ATP standards of known concentrations were used to convert relative luminescent units readings to ATP concentrations. Neutral red uptake analysis was performed using a commercially available neutral red kit (Sigma). The neutral red dye was added to the incubation media for 3 h at 37 8C, 5% CO2. The dye was fixed and solublized, and the amount of neutral red taken up by the hepatocytes was measured photometrically. MTT metabolism analysis was performed using a commercially available MTT reagent (Sigma). The reagent was added to the incubation media at the end of the troglitazone or rosiglitazone treatment and incubated for an additional 3 h at 37 8C, 5% CO2. The formazan crystals formed were dissolved in acidified isopropanol and measured photometrically following overnight incubation at 4 8C. Alamar Blue was purchased from Bio Source International (Camarillo, CA) and added to the incubation media for 3 h at 37 8C, 5% CO2, and its reduction was measured fluorometrically. GSH content was measured using a modified version of previously published methods ([16]) based on the reaction of GSH with o -pthaldehyde (Sigma). Fluorescent emission values for the GSH assay were converted to micromolar GSH concentrations using a standard curve based on the incubation of opthaldehyde with known concentrations of glutathione (Sigma).
2.3. Data analysis EC50 values for troglitazone and rosiglitazone, and associated statistical analyses (Student’s t -test) were generated using Graph Pad (Prism Computer Software, San Diego, CA). Least squares regression and other statistical analyses between donor properties and EC50 values were investigated using Microsoft Excel.
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
65
3. Results 3.1. Troglitazone and rosiglitazone cytotoxicity Troglitazone or rosiglitazone cytotoxicity was evaluated in cryopreserved human hepatocytes from 37 donors using ATP content as an endpoint following 2 h treatment with either drug (Fig. 1). Troglitazone was found to be cytotoxic (EC50 B/ 100 mM) in 86% (32 of 37) of the donors screened, while rosiglitazone appeared to be cytotoxic in 19% (7 of 37) of the donors (Table 1). In one case, the EC50 value fell below 50 mM for troglitazone (Lot 109, Table 1). Cytotoxicity of both drugs was also measured in freshly isolated human hepatocytes following a 24 h incubation using ATP content, MTT metabolism, GSH depletion, Alamar Blue metabolism and neutral red uptake as endpoints (Fig. 2). Troglitazone was found to be cytotoxic (EC50 B/100 mM) to fresh hepatocytes by all endpoints except GSH depletion, while rosigliatazone was not found to be cytotoxic by any of the endpoints considered. While some depletion of GSH by both drugs was observed in fresh hepatocytes, neither drug caused GSH levels to fall below 70% of vehicle control. 3.2. Correlation of age, sex, and environmental factors with sensitivity to troglitazone or rosiglitazone The demographic information for the 37 cryopreserved hepatocyte donors used are shown in Table 1. No apparent correlation was observed between EC50 values of troglitazone (Fig. 3) or rosiglitazone (Fig. 4) and age, sex, tobacco use or alcohol consumption.
4. Discussion IDT (adverse drug reactions which occur at low frequency in the human population) is a major challenge in drug development. Current pre-clinical and clinical safety trials appear to be inadequate predictors of IDT. One approach to understanding IDT is to elucidate the key properties responsible for the observed toxicity with the aim of identifying susceptible human subpopulations. Troglitazone was removed from the market as a result of its association with IDT in the clinical setting ([1]). Rosiglitazone, which has been associated with IDT in only two isolated cases worldwide, is currently on the market as a replacement drug of the same class ([8]). In the clinical setting, the differences in toxicity may due in part to the fact that troglitazone is administered at a relatively high dose compared to rosiglitazone. The clinical dose difference between these two drugs may relate to why the EC50 values in Table 1 for rosiglitazone are no more than two times the values for troglitazone in most donors, since the same doses of each drug were used in vitro. We evaluated troglitazone and rosiglitazone cytotoxicity in cryopreserved human hepatocytes from multiple donors as part of our research program to develop approaches to predict IDT. One purpose of this study was to evaluate and compare
Lot
Trog Race
EC50
EC50
Age
Sex
Weight (kg)
Smoker Alcohol use
Medications/drugs
Disease history
Occasional marijuana as teen, hydrocordone No Marijuana Agent orange exposure
Childhood asthma No No Heart disease
Occasional marijuana Xanax (not prescribed) No
Hypertension, 15 /20 yrs Childhood asthma No
N/A
Hypertension
No
Anemia 1 yr ago
No
No
N/A N/A Nitropaste
N/A Chronic cough, kidney infection Emphysema, mild arthritis, myocardial infarction (15 /17 yrs ago) N/A Heart disease, hypertension Hemia, hypertension NIDDM (10 yrs), hypertension, high cholesterol transitional blood cancer Hypertension N/A N/A
59
101.7
92.6 Caucasian 33
Male
75
Yes
Social
61 62 66
62.6 73.1 146.2
63.7 Caucasian 38 63.6 Caucasian 46 95.1 Caucasian 48
Male Male Male
95 102 96
Yes No No
70 71 75
95.3 103.2 188.6
Male Male Male
113 91 10
Yes No No
83
128.4
52 Caucasian 57 83.4 Hispanic 23 92.5 Caucasian 15 months 92.1 Black 56
Male
89
No
86
135.9
76.3 Caucasian 73
Female 51
No
88
142.7
95.6 Caucasian 84
Female 62
No
Female 85 Female N/A Female 60
Yes Yes Yes
Male Male Male Male
170 111 64 88
No Yes No No
Daily Daily Nondrinker Social Social Nondrinker Nondrinker Nondrinker Nondrinker N/A Social Nondrinker N/A Social Daily Social
Female 52 Female 50 Female 100
Yes Yes No
Social Social Social
89 90 91
128.4 92.1 Caucasian 44 162.2 78 Caucasian 51 /500 139.4 Caucasian 72
95 105 108 109
138.8 64.9 87.4 67.1 106.3 105 128.6 20.6
111 113 114
88.8 132.2 88.1
Caucasian Caucasian Asian Caucasian
43 59 59 69
55.7 Black 59 91.4 Caucasian 61 82.7 Caucasian 47
Occasional marijuana Cardizem, zocor N/A N/A Diazide No Opiates
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Rosi
66
Table 1 Background and demographic information on cryopreserved human hepatocyte donors screened for sensitivity to troglitazone and rosiglitazone with EC50 values from ATP assays
Table 1 (Continued ) Lot
Rosi
Trog Race
EC50
EC50
Age
Weight (kg)
Smoker Alcohol use
Medications/drugs
Disease history
57 57 177 N/A 15
Yes Yes No N/A No
N/A N/A Some marijuana N/A No
N/A N/A Hypertension (untreated) N/A No
118 119 122 129 130
233.8 142.8 Caucasian 48 191.4 90.7 Caucasian 48 124.4 71.9 Hispanic 42 196.4 92.2 N/A N/A 111.7 77.8 Caucasian 2
Female Female Male N/A Female
133
125.3
81.5 Hispanic
Female 79
Yes
Daily Daily Daily N/A Nondrinker Social
CNV
133.8
93.2 Caucasian 38
Male
103
Yes
Social
DRL EFA
103.8 78.9
78.1 Caucasian 44 63.2 Caucasian 16
Male Male
100 73
Yes No
ETR
117.9
85.6 Caucasian 51
Female 74
No
EVY
105.8
69.3 Caucasian 63
Male
No
Social Nondrinker Nondrinker Daily
GNG HRK
105.4 85.3 Caucasian 45 124.5 115.2 Caucasian 66
59
124
Female 84 Female 90
Yes No
KMI 103.7 101.3 Caucasian 45 MQF 123.3 57.8 Caucasian 49 MYO /500 62.6 Caucasian 59
Male 110 Female 104 Female 90
No No No
TVC
Female 55
Yes
74.4
91.7 Caucasian 58
Social Nondrinker Social Social Nondrinker Nondrinker
Provitol, quanarin, monopril,- Depression celexia, flovent, deserel Catenolol Hypertension (B/5 yrs), hepatitis C in late 1980s No N/A No No No
No
Cumadin, prinavil, lassix,digitech No 25 yrs (for hypertension)
N/A No Hypertension
No N/A No
Hypertension N/A Adult onset diabetes
Provacol, claritan, plavix, climare, lopressor
Vascular disease
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Sex
67
68
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 2. Dose response curves to troglitazone (triangles) or rosiglitazone (squares) for freshly isolated primary hepatocytes from a single donor as evaluated by ATP production (panel A), MTT metabolism (panel B), GSH depletion (panel C), Alamar Blue metabolism (panel D) or Neutral Red uptake (panel E).
the potential hepatotoxic effects of these two closely related drugs among a large panel of individuals with diverse demographic backgrounds using an in vitro approach. This was done in an effort to assess the sensitivity of our pre-clinical in vitro model system by determining the cytotoxic effects of a known causative agent of IDT in comparison with a structurally similar compound not associated with IDT. Our results correlate well with the clinical findings associated with the thiazolidinediones. Troglitazone was found to be cytotoxic (EC50 B/100 mM) at a rate of 86% in our in vitro study. Our results suggest that cryopreserved human hepatocytes provide a good pre-clinical in vitro model system for the evaluation of potential causation of IDT in the early stages of drug development. In addition, the use of freshly isolated hepatocytes in such studies allows for toxicity evaluation
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
69
Fig. 3. Correlation of troglitazone hepatotoxicity with donor age (panel A), sex (panel B), smoking status (panel C) or level of alcohol consumption (panel D).
following extended incubation periods. However, the availability of fresh tissue imposes limits on the number of donors that can be included in such studies. Finally, the use of multiple endpoints may provide insights into the mechanisms of IDT and help to eliminate experimental artifacts specific to a single endpoint. A secondary purpose of the study was to clarify relationships between troglitazone or rosiglitazone cytotoxicity and donor age, sex, alcohol consumption, and tobacco use, while comparing the observed in vitro hepatotoxic effects of these two related drugs. We observed no apparent correlation between cytotoxicity and the donor properties considered for either drug. It is important to note, however, that the availability of demographic data specific to each donor empowers researches with the ability to establish relationships between specific donor attributes and IDT. In summary, our results provide evidence for the usefulness of cryopreserved hepatocytes in the evaluation of drug toxicity in vitro. The results of the study relate well to the clinical findings that troglitazone is more toxic than rosiglitazone to the liver. However, a significant percentage of individuals were also found to be sensitive to rosiglitazone in our study, suggesting the need for further clinical evaluation or, at
70
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 4. Correlation of rosiglitazone hepatotoxicity with donor age (panel A), sex (panel B), smoking status (panel C) or level of alcohol consumption (panel D).
a minimum, the close and careful monitoring of liver function in patients receiving this compound for the treatment of type 2 diabetes.
Acknowledgements Presley Mozone, In Vitro Technologies, Inc., for assistance with hepatocyte handling and cryostorage. Danielle F. Hill, In Vitro Technologies, Inc., for assistance with hepatocyte handling and cryostorage. Blaise Considine, In Vitro Technologies, Inc., for assistance with manuscript preparation.
References [1] D.L. Brown, D. Brillon, New directions in type 2 diabetes mellitus: an update of current oral antidiabetic therapy, J. Natl. Med. Assoc. 91 (7) (1999) 389 /395. [2] C. Kausch, J. Krutzfeldt, A. Witke, A. Rettig, O. Bachmann, K. Rett, S. Matthaei, F. Machicai, H.U. Haring, M. Stumvoll, Effects of troglitazone on cellular differentiation, insulin signaling, and glucose metabolism in cultured human skeletal muscle cells, Biochem. Biophys. Res. Com. 280 (2001) 664 / 674.
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
71
[3] A.R. Saltiel, J.M. Olefsky, Thiazolidinediones in the treatment of insulin resistance and type II diabetes, Diabetes 45 (1996) 1661 /1669. [4] B.M. Spiegelman, Perspectives in diabetes. PPAR-g: adipogenic regulator and thiazolidinedione receptor, Diabetes 47 (1998) 507 /514. [5] P. Biswas, L.V. Wilton, S.A. Shakir, Troglitazone and liver function abnormalities, Drug Saf. 24 (2) (2001) 149 /154. [6] R. Donnelly, FDA Reviews Troglitazone, Diab. Obes. Metab. 1 (1999) 65 /66. [7] CSM/MCA, Troglitazone (Romozin) withdrawn, Curr. Probl. Pharmacovigil 23 (1997) 12. [8] A.J. Scheen, Thiazolidinediones and liver toxicity, Diabetes Metab. 27 (3) (2001) 305 /313. [9] L.M. Forman, D.A. Simmons, R.H. Diamond, Hepatic failure in a patient taking rosiglitazone, Ann. Intern. Med. 132 (2000) 118 /121. [10] J. Al-Salman, H. Arjomand, et al., Hepatocellular injury in a patient receiving rosiglitazone, Ann. Intern. Med. 132 (2000) 121 /124. [11] Y. Toyoda, A. Tsuchida, E. Iwami, I. Miwa, Toxic effect of troglitazone on cultured rat hepatocytes (2001), Life Sci. 68 (2001) 1867 /1876. [12] C. Funk, C. Ponelle, G. Scheuermann, M. Pantze, Cholestatic potential of troglitazone as a possible factor contributing to troglitazone-induced hepatotoxicity; in vivo and in vitro interaction at the canalicular bile salt export pump (bsep) in the rat, Mol. Pharmacol. 59 (2001) 627 /635. [13] L.J. Loretz, A.P. Li, M.W. Flye, A.G.E. Wilson, Optimization of cryopreservation procedures for rat and human hepatocytes, Xenobiotica 19 (1989) 489 /498. [14] A.P. Li, M.A. Roque, D.J. Beck, D.L. Kaminski, Isolation and culturing of hepatocytes from human livers, J. Tiss. Cult. Meth. 14 (1992) 139 /146. [15] A.P. Li, C. Lu, J.A. Brendt, C. Pham, A. Fackett, C.E. Ruegg, P.A. Silber, Cryopreserved human hepatocytes: characterization of drug-metabolizing enzyme activities and applications in higher throughput screening assays for hepatotoxicity, metabolic stability, and drug /drug interaction potential, Chem. Biol. Interact. 121 (1999) 17 /35. [16] A.P. Senft, T.P. Dalton, H.G. Shertzer, Determining glutathione and glutathione disulfide using the fluorescence probe o -pthaldehyde, Anal. Biochem. 280 (1) (2000) 80 /86.
Differential in vitro hepatotoxicity of troglitazone and rosiglitazone among cryopreserved human hepatocytes from 37 donors Scott Lloyd *, Michael J. Hayden, Yumiko Sakai, Andrew Fackett, Paul M. Silber, Nicola J. Hewitt, Albert P. Li In Vitro Technologies, Inc., 1450 South Rolling Road, Baltimore, MD 21227, USA
Abstract We report here our studies on troglitazone and rosiglitazone cytotoxicity in human hepatocytes isolated from multiple donors to investigate factors responsible for individual differences in sensitivity to the known hepatotoxicity of these antidiabetic drugs. Using cellular adenosine triphosphate (ATP) content as an endpoint, cytotoxicity of both drugs was evaluated in cryopreserved human hepatocytes from 37 donors. We confirmed reports of others that troglitazone was cytotoxic to human hepatocytes using cellular ATP content as an endpoint. In addition, we found that rosiglitazone, although less toxic in the study population, was cytotoxic to hepatocytes in some donors (EC50 B/100 mM). ATP content, 3-[4,5dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) metabolism, depletion of intracellular glutathione, Alamar Blue metabolism, and neutral red uptake were used as endpoints in a single donor study using freshly isolated human hepatocytes. Troglitazone appeared to be more toxic than rosiglitazone by all endpoints. From the demographic data provided to us for each donor, we were able to establish no direct correlation between cytotoxicity (expressed as EC50 values) and age, sex, smoking status, or alcohol consumption. We conclude that troglitazone and rosiglitazone are differentially toxic to human hepatocytes, and that toxicity may be independent of age, sex, tobacco use, and alcohol use. # 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Adenosine triphosphate; Hepatocytes; Troglitazone
Support by In Vitro Technologies, Inc., Baltimore, Maryland. This work was presented, in part, at the second International Symposium on Mechanisms, Models and Predictions of Idiosyncratic Drug Toxicity, Atlantic City, New Jersey, June 11 /13, 2001. * Corresponding author. Tel.: /1-410-455-1242; fax: /1-410-455-1245 E-mail address: [email protected] (S. Lloyd).
0009-2797/02/$ - see front matter # 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 0 0 9 - 2 7 9 7 ( 0 2 ) 0 0 0 5 4 - 6
58
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
1. Introduction The thiazolidinediones are a relatively new class of drugs used to treat (type 2) non-insulin dependent diabetes mellitus, or NIDDM ([1]). This novel class of antidiabetic agents has proven useful as either a monotherapy, or in combination therapy with insulin, metformin or sulfonylureas in the treatment of NIDDM ([1]). The thiazolidinediones are selective agonists of the peroxisome proliferator-activated receptors (PPARs) and have been shown to lower insulin resistance in both diabetic and non-diabetic individuals ([2,3]). The PPARs, especially PPARg, are cell surface receptors and transcription factors present in adipocytes, skeletal muscle, and liver
Fig. 1. Dose response curves to troglitazone (triangles) and rosiglitazone (squares) for cryopreserved hepatocytes from 37 different donors (panels A through KK).
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
59
Fig. 1 (Continued)
tissue, and are key elements in the regulation and metabolism of glucose in response to insulin ([4]). Many patients suffering from NIDDM lack glycemic control due to non-responsiveness of PPARs, making the use of thiazolidinediones an attractive therapeutic approach. Troglitazone was introduced to the market in 1997 and was the first thiazolidinedione to be approved for use in the treatment of NIDDM ([1]). However, during its first 2 years on the market, a number of cases were reported that associated troglitazone with acute liver toxicity and failure, leading to the death of patients using the drug or necessitating liver transplant in some cases ([5,6]). Subsequent clinical investigations lead to its withdrawal from the UK market in December 1997 and to its withdrawal from the US market in March of 2000 based on data that associated troglitazone with cases of serious idiosyncratic liver toxicity ([7]).
60
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
Rosiglitazone, a relatively non-toxic alternative to troglitazone, was introduced to the market in 1999 ([1]). While the incidence of significant hepatotoxicity caused by rosiglitazone has not been observed in controlled clinical trials, two cases have been reported which implicate rosiglitazone as a causative agent of idiosyncratic drug toxicity (IDT) as well ([8 /10]). Several studies have proposed a variety of mechanisms of troglitazone cytotoxicity, including parent compound toxicity in human hepatocytes, the formation of reactive intermediates and glutathione (GSH) conjugates in human liver microsomes, and both cholestatic and apoptotic potentials in cultured rat hepatocytes ([11,12]). Mechanisms of hepatocellular injury caused by rosiglitazone have not been well studied, likely because the causal relationship between the drug and hepatotoxicity remains uncertain. This study used a panel of cryopreserved human hepatocytes isolated from demographically diverse donors,
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
61
Fig. 1 (Continued)
and freshly isolated hepatocytes from a single donor, to demonstrate dose-dependent troglitazone toxicity and to compare individual sensitivities to troglitazone and rosiglitazone on the basis of age, sex, alcohol or tobacco use.
2. Materials and methods 2.1. Incubation of hepatocytes with troglitazone and rosiglitazone Cryopreserved hepatocytes from 37 individual donors, along with freshly isolated hepatocytes from a single donor, were used in this study. The cells were isolated,
62
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
cryopreserved, and thawed using previously published procedures ([13 /15]). Troglitazone and rosiglitazone were obtained from Parke-Davis (Ann Arbor, MI) in powder form and prepared as a 2000 /stock solution in dimethyl sulphoxide (DMSO; Sigma, St. Louis, MO), then diluted to a 2 /concentrated dosing solution in Dulbecco’s Modified Eagle Medium (Sigma). Cytotoxicity assays were initiated by the addition of an equal volume of cryopreserved hepatocyte suspension to the media containing troglitazone or rosiglitazone. The final troglitazone and rosiglitazone concentrations used in the incubations were 10, 20, 50, 100 and 200 mM. The final DMSO concentration was 0.1% (v/v). In a single related experiment, freshly isolated hepatocytes were incubated overnight on collagen (Cohesion, Palo Alto, CA) coated 96 well plates. The plating media was then replaced with incubation
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 1 (Continued)
63
64
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
media containing 10, 20, 50, 100 or 200 mM troglitazone or rosiglitazone for a 24-h incubation period. Incubations of hepatocytes with drug and control solutions were carried out in a cell culture incubator maintained at 37 8C, 5% CO2 in saturating humidity. The treatment duration was 2 h for cryopreserved hepatocytes.
2.2. Cell viability analysis ATP content analysis was performed using the ATP Lite-M kit from Packard Marketing (Groningen, The Netherlands). Briefly, the hepatocytes were lysed after incubation followed by the addition of luciferin and luciferase. After a 10-min dark adaptation, luciferin luminescence was quantified using a Wallac Victor 2 Microplate Reader (Perkin Elmer, Gaithersburg, MD). ATP standards of known concentrations were used to convert relative luminescent units readings to ATP concentrations. Neutral red uptake analysis was performed using a commercially available neutral red kit (Sigma). The neutral red dye was added to the incubation media for 3 h at 37 8C, 5% CO2. The dye was fixed and solublized, and the amount of neutral red taken up by the hepatocytes was measured photometrically. MTT metabolism analysis was performed using a commercially available MTT reagent (Sigma). The reagent was added to the incubation media at the end of the troglitazone or rosiglitazone treatment and incubated for an additional 3 h at 37 8C, 5% CO2. The formazan crystals formed were dissolved in acidified isopropanol and measured photometrically following overnight incubation at 4 8C. Alamar Blue was purchased from Bio Source International (Camarillo, CA) and added to the incubation media for 3 h at 37 8C, 5% CO2, and its reduction was measured fluorometrically. GSH content was measured using a modified version of previously published methods ([16]) based on the reaction of GSH with o -pthaldehyde (Sigma). Fluorescent emission values for the GSH assay were converted to micromolar GSH concentrations using a standard curve based on the incubation of opthaldehyde with known concentrations of glutathione (Sigma).
2.3. Data analysis EC50 values for troglitazone and rosiglitazone, and associated statistical analyses (Student’s t -test) were generated using Graph Pad (Prism Computer Software, San Diego, CA). Least squares regression and other statistical analyses between donor properties and EC50 values were investigated using Microsoft Excel.
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
65
3. Results 3.1. Troglitazone and rosiglitazone cytotoxicity Troglitazone or rosiglitazone cytotoxicity was evaluated in cryopreserved human hepatocytes from 37 donors using ATP content as an endpoint following 2 h treatment with either drug (Fig. 1). Troglitazone was found to be cytotoxic (EC50 B/ 100 mM) in 86% (32 of 37) of the donors screened, while rosiglitazone appeared to be cytotoxic in 19% (7 of 37) of the donors (Table 1). In one case, the EC50 value fell below 50 mM for troglitazone (Lot 109, Table 1). Cytotoxicity of both drugs was also measured in freshly isolated human hepatocytes following a 24 h incubation using ATP content, MTT metabolism, GSH depletion, Alamar Blue metabolism and neutral red uptake as endpoints (Fig. 2). Troglitazone was found to be cytotoxic (EC50 B/100 mM) to fresh hepatocytes by all endpoints except GSH depletion, while rosigliatazone was not found to be cytotoxic by any of the endpoints considered. While some depletion of GSH by both drugs was observed in fresh hepatocytes, neither drug caused GSH levels to fall below 70% of vehicle control. 3.2. Correlation of age, sex, and environmental factors with sensitivity to troglitazone or rosiglitazone The demographic information for the 37 cryopreserved hepatocyte donors used are shown in Table 1. No apparent correlation was observed between EC50 values of troglitazone (Fig. 3) or rosiglitazone (Fig. 4) and age, sex, tobacco use or alcohol consumption.
4. Discussion IDT (adverse drug reactions which occur at low frequency in the human population) is a major challenge in drug development. Current pre-clinical and clinical safety trials appear to be inadequate predictors of IDT. One approach to understanding IDT is to elucidate the key properties responsible for the observed toxicity with the aim of identifying susceptible human subpopulations. Troglitazone was removed from the market as a result of its association with IDT in the clinical setting ([1]). Rosiglitazone, which has been associated with IDT in only two isolated cases worldwide, is currently on the market as a replacement drug of the same class ([8]). In the clinical setting, the differences in toxicity may due in part to the fact that troglitazone is administered at a relatively high dose compared to rosiglitazone. The clinical dose difference between these two drugs may relate to why the EC50 values in Table 1 for rosiglitazone are no more than two times the values for troglitazone in most donors, since the same doses of each drug were used in vitro. We evaluated troglitazone and rosiglitazone cytotoxicity in cryopreserved human hepatocytes from multiple donors as part of our research program to develop approaches to predict IDT. One purpose of this study was to evaluate and compare
Lot
Trog Race
EC50
EC50
Age
Sex
Weight (kg)
Smoker Alcohol use
Medications/drugs
Disease history
Occasional marijuana as teen, hydrocordone No Marijuana Agent orange exposure
Childhood asthma No No Heart disease
Occasional marijuana Xanax (not prescribed) No
Hypertension, 15 /20 yrs Childhood asthma No
N/A
Hypertension
No
Anemia 1 yr ago
No
No
N/A N/A Nitropaste
N/A Chronic cough, kidney infection Emphysema, mild arthritis, myocardial infarction (15 /17 yrs ago) N/A Heart disease, hypertension Hemia, hypertension NIDDM (10 yrs), hypertension, high cholesterol transitional blood cancer Hypertension N/A N/A
59
101.7
92.6 Caucasian 33
Male
75
Yes
Social
61 62 66
62.6 73.1 146.2
63.7 Caucasian 38 63.6 Caucasian 46 95.1 Caucasian 48
Male Male Male
95 102 96
Yes No No
70 71 75
95.3 103.2 188.6
Male Male Male
113 91 10
Yes No No
83
128.4
52 Caucasian 57 83.4 Hispanic 23 92.5 Caucasian 15 months 92.1 Black 56
Male
89
No
86
135.9
76.3 Caucasian 73
Female 51
No
88
142.7
95.6 Caucasian 84
Female 62
No
Female 85 Female N/A Female 60
Yes Yes Yes
Male Male Male Male
170 111 64 88
No Yes No No
Daily Daily Nondrinker Social Social Nondrinker Nondrinker Nondrinker Nondrinker N/A Social Nondrinker N/A Social Daily Social
Female 52 Female 50 Female 100
Yes Yes No
Social Social Social
89 90 91
128.4 92.1 Caucasian 44 162.2 78 Caucasian 51 /500 139.4 Caucasian 72
95 105 108 109
138.8 64.9 87.4 67.1 106.3 105 128.6 20.6
111 113 114
88.8 132.2 88.1
Caucasian Caucasian Asian Caucasian
43 59 59 69
55.7 Black 59 91.4 Caucasian 61 82.7 Caucasian 47
Occasional marijuana Cardizem, zocor N/A N/A Diazide No Opiates
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Rosi
66
Table 1 Background and demographic information on cryopreserved human hepatocyte donors screened for sensitivity to troglitazone and rosiglitazone with EC50 values from ATP assays
Table 1 (Continued ) Lot
Rosi
Trog Race
EC50
EC50
Age
Weight (kg)
Smoker Alcohol use
Medications/drugs
Disease history
57 57 177 N/A 15
Yes Yes No N/A No
N/A N/A Some marijuana N/A No
N/A N/A Hypertension (untreated) N/A No
118 119 122 129 130
233.8 142.8 Caucasian 48 191.4 90.7 Caucasian 48 124.4 71.9 Hispanic 42 196.4 92.2 N/A N/A 111.7 77.8 Caucasian 2
Female Female Male N/A Female
133
125.3
81.5 Hispanic
Female 79
Yes
Daily Daily Daily N/A Nondrinker Social
CNV
133.8
93.2 Caucasian 38
Male
103
Yes
Social
DRL EFA
103.8 78.9
78.1 Caucasian 44 63.2 Caucasian 16
Male Male
100 73
Yes No
ETR
117.9
85.6 Caucasian 51
Female 74
No
EVY
105.8
69.3 Caucasian 63
Male
No
Social Nondrinker Nondrinker Daily
GNG HRK
105.4 85.3 Caucasian 45 124.5 115.2 Caucasian 66
59
124
Female 84 Female 90
Yes No
KMI 103.7 101.3 Caucasian 45 MQF 123.3 57.8 Caucasian 49 MYO /500 62.6 Caucasian 59
Male 110 Female 104 Female 90
No No No
TVC
Female 55
Yes
74.4
91.7 Caucasian 58
Social Nondrinker Social Social Nondrinker Nondrinker
Provitol, quanarin, monopril,- Depression celexia, flovent, deserel Catenolol Hypertension (B/5 yrs), hepatitis C in late 1980s No N/A No No No
No
Cumadin, prinavil, lassix,digitech No 25 yrs (for hypertension)
N/A No Hypertension
No N/A No
Hypertension N/A Adult onset diabetes
Provacol, claritan, plavix, climare, lopressor
Vascular disease
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Sex
67
68
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 2. Dose response curves to troglitazone (triangles) or rosiglitazone (squares) for freshly isolated primary hepatocytes from a single donor as evaluated by ATP production (panel A), MTT metabolism (panel B), GSH depletion (panel C), Alamar Blue metabolism (panel D) or Neutral Red uptake (panel E).
the potential hepatotoxic effects of these two closely related drugs among a large panel of individuals with diverse demographic backgrounds using an in vitro approach. This was done in an effort to assess the sensitivity of our pre-clinical in vitro model system by determining the cytotoxic effects of a known causative agent of IDT in comparison with a structurally similar compound not associated with IDT. Our results correlate well with the clinical findings associated with the thiazolidinediones. Troglitazone was found to be cytotoxic (EC50 B/100 mM) at a rate of 86% in our in vitro study. Our results suggest that cryopreserved human hepatocytes provide a good pre-clinical in vitro model system for the evaluation of potential causation of IDT in the early stages of drug development. In addition, the use of freshly isolated hepatocytes in such studies allows for toxicity evaluation
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
69
Fig. 3. Correlation of troglitazone hepatotoxicity with donor age (panel A), sex (panel B), smoking status (panel C) or level of alcohol consumption (panel D).
following extended incubation periods. However, the availability of fresh tissue imposes limits on the number of donors that can be included in such studies. Finally, the use of multiple endpoints may provide insights into the mechanisms of IDT and help to eliminate experimental artifacts specific to a single endpoint. A secondary purpose of the study was to clarify relationships between troglitazone or rosiglitazone cytotoxicity and donor age, sex, alcohol consumption, and tobacco use, while comparing the observed in vitro hepatotoxic effects of these two related drugs. We observed no apparent correlation between cytotoxicity and the donor properties considered for either drug. It is important to note, however, that the availability of demographic data specific to each donor empowers researches with the ability to establish relationships between specific donor attributes and IDT. In summary, our results provide evidence for the usefulness of cryopreserved hepatocytes in the evaluation of drug toxicity in vitro. The results of the study relate well to the clinical findings that troglitazone is more toxic than rosiglitazone to the liver. However, a significant percentage of individuals were also found to be sensitive to rosiglitazone in our study, suggesting the need for further clinical evaluation or, at
70
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
Fig. 4. Correlation of rosiglitazone hepatotoxicity with donor age (panel A), sex (panel B), smoking status (panel C) or level of alcohol consumption (panel D).
a minimum, the close and careful monitoring of liver function in patients receiving this compound for the treatment of type 2 diabetes.
Acknowledgements Presley Mozone, In Vitro Technologies, Inc., for assistance with hepatocyte handling and cryostorage. Danielle F. Hill, In Vitro Technologies, Inc., for assistance with hepatocyte handling and cryostorage. Blaise Considine, In Vitro Technologies, Inc., for assistance with manuscript preparation.
References [1] D.L. Brown, D. Brillon, New directions in type 2 diabetes mellitus: an update of current oral antidiabetic therapy, J. Natl. Med. Assoc. 91 (7) (1999) 389 /395. [2] C. Kausch, J. Krutzfeldt, A. Witke, A. Rettig, O. Bachmann, K. Rett, S. Matthaei, F. Machicai, H.U. Haring, M. Stumvoll, Effects of troglitazone on cellular differentiation, insulin signaling, and glucose metabolism in cultured human skeletal muscle cells, Biochem. Biophys. Res. Com. 280 (2001) 664 / 674.
S. Lloyd et al. / Chemico-Biological Interactions 142 (2002) 57 /71
71
[3] A.R. Saltiel, J.M. Olefsky, Thiazolidinediones in the treatment of insulin resistance and type II diabetes, Diabetes 45 (1996) 1661 /1669. [4] B.M. Spiegelman, Perspectives in diabetes. PPAR-g: adipogenic regulator and thiazolidinedione receptor, Diabetes 47 (1998) 507 /514. [5] P. Biswas, L.V. Wilton, S.A. Shakir, Troglitazone and liver function abnormalities, Drug Saf. 24 (2) (2001) 149 /154. [6] R. Donnelly, FDA Reviews Troglitazone, Diab. Obes. Metab. 1 (1999) 65 /66. [7] CSM/MCA, Troglitazone (Romozin) withdrawn, Curr. Probl. Pharmacovigil 23 (1997) 12. [8] A.J. Scheen, Thiazolidinediones and liver toxicity, Diabetes Metab. 27 (3) (2001) 305 /313. [9] L.M. Forman, D.A. Simmons, R.H. Diamond, Hepatic failure in a patient taking rosiglitazone, Ann. Intern. Med. 132 (2000) 118 /121. [10] J. Al-Salman, H. Arjomand, et al., Hepatocellular injury in a patient receiving rosiglitazone, Ann. Intern. Med. 132 (2000) 121 /124. [11] Y. Toyoda, A. Tsuchida, E. Iwami, I. Miwa, Toxic effect of troglitazone on cultured rat hepatocytes (2001), Life Sci. 68 (2001) 1867 /1876. [12] C. Funk, C. Ponelle, G. Scheuermann, M. Pantze, Cholestatic potential of troglitazone as a possible factor contributing to troglitazone-induced hepatotoxicity; in vivo and in vitro interaction at the canalicular bile salt export pump (bsep) in the rat, Mol. Pharmacol. 59 (2001) 627 /635. [13] L.J. Loretz, A.P. Li, M.W. Flye, A.G.E. Wilson, Optimization of cryopreservation procedures for rat and human hepatocytes, Xenobiotica 19 (1989) 489 /498. [14] A.P. Li, M.A. Roque, D.J. Beck, D.L. Kaminski, Isolation and culturing of hepatocytes from human livers, J. Tiss. Cult. Meth. 14 (1992) 139 /146. [15] A.P. Li, C. Lu, J.A. Brendt, C. Pham, A. Fackett, C.E. Ruegg, P.A. Silber, Cryopreserved human hepatocytes: characterization of drug-metabolizing enzyme activities and applications in higher throughput screening assays for hepatotoxicity, metabolic stability, and drug /drug interaction potential, Chem. Biol. Interact. 121 (1999) 17 /35. [16] A.P. Senft, T.P. Dalton, H.G. Shertzer, Determining glutathione and glutathione disulfide using the fluorescence probe o -pthaldehyde, Anal. Biochem. 280 (1) (2000) 80 /86.